def control(v, t, x):
        goal = (6, 6)
        goal_heading = atan2(goal[1] - x[1], goal[0] - x[0])
        d_heading = base.angdiff(goal_heading, x[2])
        v.stopif(base.norm(x[0:2] - goal) < 0.1)

        return (1, d_heading)
Esempio n. 2
0
    def demand(self):
        """
        Compute speed and heading for random waypoint
            %
            % [SPEED,STEER] = R.demand() is the speed and steer angle to
            % drive the vehicle toward the next waypoint.  When the vehicle is
            % within R.dtresh a new waypoint is chosen.
            %
            % See also Vehicle."""

        if self._goal is None:
            self._new_goal()

        # if nearly at goal point, choose the next one
        d = np.linalg.norm(self._veh._x[0:2] - self._goal)
        if d < self._dthresh:
            self._new_goal()
        # elif d > 2 * self._d_prev:
        #     self.choose_goal()
        # self._d_prev = d

        speed = self._speed

        goal_heading = atan2(self._goal[1] - self._veh._x[1],
                             self._goal[0] - self._veh._x[0])
        delta_heading = base.angdiff(goal_heading, self._veh._x[2])

        return np.r_[speed, self._headinggain * delta_heading]
    def h(self, xv, jf=None):
        """
        Landmark range and bearing
        %
        # Z = self.h(X, K) is a sensor observation (1x2), range and bearing, from vehicle at 
        # pose X (1x3) to the K'th landmark.
        %
        # Z = self.h(X, P) as above but compute range and bearing to a landmark at coordinate P.
        %
        # Z = self.h(X) as above but computes range and bearing to all
        # map featureself.  Z has one row per landmark.
        %
        # Notes::
        # - Noise with covariance W (propertyW) is added to each row of Z.
        # - Supports vectorized operation where XV (Nx3) and Z (Nx2).
        # - The landmark is assumed visible, field of view and range liits are not
        #   applied.
        %
        # See also RangeBearingSensor.reading, RangeBearingSensor.Hx, RangeBearingSensor.Hw, RangeBearingSensor.Hp.
        """
        # get the landmarks, one per row
        if jf is None:
            # self.h(XV)   all landmarks
            dx = self.map.x - xv[0]
            dy = self.map.y - xv[1]
        elif base.isinteger(jf):
            # self.h(XV, JF)
            xlm = self.map.landmark(jf)
            dx = xlm[0] - xv[0]
            dy = xlm[1] - xv[1]
        else:
            # self.h(XV, XF)
            xlm = base.getvector(jf, 2)
            dx = xlm[0] - xv[0]
            dy = xlm[1] - xv[1]

        # compute range and bearing (Vectorized code)

        z = np.c_[np.sqrt(dx**2 + dy**2),
                  base.angdiff(np.arctan2(dy, dx),
                               xv[2])]  # range & bearing as columns

        # add noise with covariance W
        z += self._rand.normal(size=z.shape) @ sp.linalg.sqrtm(self._W)

        if z.shape[0] == 1:
            return z[0]
        else:
            return z
        def ik3(robot, T, config='lun'):

            config = self.config_validate(config, ('lr', 'ud', 'nf'))

            # solve for the first three joints

            a2 = robot.links[1].a
            a3 = robot.links[2].a
            d1 = robot.links[0].d
            d3 = robot.links[2].d
            d4 = robot.links[3].d

            # The following parameters are extracted from the Homogeneous
            # Transformation as defined in equation 1, p. 34

            Px, Py, Pz = T.t
            Pz -= d1  # offset the pedestal height
            theta = np.zeros((3, ))

            # Solve for theta[0]
            # r is defined in equation 38, p. 39.
            # theta[0] uses equations 40 and 41, p.39,
            # based on the configuration parameter n1

            r = np.sqrt(Px**2 + Py**2)
            if 'r' in config:
                theta[0] = np.arctan2(Py, Px) + np.arcsin(d3 / r)
            elif 'l' in config:
                theta[0] = np.arctan2(Py, Px) + np.pi - np.arcsin(d3 / r)
            else:
                raise ValueError('bad configuration string')

            # Solve for theta[1]
            # V114 is defined in equation 43, p.39.
            # r is defined in equation 47, p.39.
            # Psi is defined in equation 49, p.40.
            # theta[1] uses equations 50 and 51, p.40, based on the
            # configuration parameter n2
            if 'u' in config:
                n2 = 1
            elif 'd' in config:
                n2 = -1
            else:
                raise ValueError('bad configuration string')

            if 'l' in config:
                n2 = -n2

            V114 = Px * np.cos(theta[0]) + Py * np.sin(theta[0])

            r = np.sqrt(V114**2 + Pz**2)

            Psi = np.arccos(
                (a2**2 - d4**2 - a3**2 + V114**2 + Pz**2) / (2.0 * a2 * r))

            if np.isnan(Psi):
                theta = None  # pragma nocover
            else:
                theta[1] = np.arctan2(Pz, V114) + n2 * Psi

                # Solve for theta[2]
                # theta[2] uses equation 57, p. 40.
                num = np.cos(theta[1]) * V114 + np.sin(theta[1]) * Pz - a2
                den = np.cos(theta[1]) * Pz - np.sin(theta[1]) * V114
                theta[2] = np.arctan2(a3, d4) - np.arctan2(num, den)

            theta = base.angdiff(theta)

            return theta
def linearize_and_solve(self):
    t0 = time.time()
    print('solving')

    # H and b are respectively the system matrix and the system vector
    H = np.zeros((self.n * 3, self.n * 3))
    b = zeros(g.n * 3, 1)
    # this loop constructs the global system by accumulating in H and b the contributions
    # of all edges (see lecture)
    fprintf('.')

    etotal = 0
    for edge in self.edges:

        e, A, B = self.linear_factors(edge)
        omega = g.edata(edge).info
        # compute the blocks of H^k

        # not quite sure whey SE3 is being transposed, what does that mean?
        b_i = -A.T @ omega @ e
        b_j = -B.T @ omega @ e
        H_ii = A.T @ omega @ A
        H_ij = A.T @ omega @ B
        H_jj = B.T @ omega @ B

        v = g.vertices(edge)

        i = v[0]
        j = v[1]

        islice = slice(i * 3, (i + 1) * 3)
        jslice = slice(j * 3, (j + 1) * 3)

        # accumulate the blocks in H and b

        H[islice, islice] += Hii
        H[jslice, jslice] += Hjj
        H[islice, jslice] += Hij
        H[jslice, islice] += Hij.T

        b[islice, 0] += b_i
        b[jslice, 0] += b_j

        etotal = etotal + np.inner(e, e)
        printf('.', end=None)

    # %note that the system (H b) is obtained only from
    # %relative constraints. H is not full rank.
    # %we solve the problem by anchoring the position of
    # %the the first vertex.
    # %this can be expressed by adding the equation
    #  deltax(1:3,1) = 0
    # which is equivalent to the following

    H[0:3, 0:3] += np.eye(3)

    SH = sp.bsr_sparse(H)
    print('.', end=None)
    deltax = sp.spsolve(SH, b)  # SH \ b
    print('.', end=None)

    # split the increments in nice 3x1 vectors and sum them up to the original matrix
    newmeans = self.coord() + np.reshape(deltax, (3, self.n))

    # normalize the angles between -PI and PI
    # for (i = 1:size(newmeans,2))
    #     s = sin(newmeans(3,i))
    #     c = cos(newmeans(3,i))
    #     newmeans(3,i) = atan2(s,c)
    newmeans[3, :] = base.angdiff(newmeans[3, :])

    dt = time.time() - t0
    print(f"done in {dt:0.2f} sec.  Total cost {etotal}")

    return newmeans, energy